Geometric Mean Maximization: An Overlooked Portfolio Approach?

Academics and practitioners usually optimize portfolios on the basis of mean and variance. They set the goal of maximizing risk-adjusted returns measured by the Sharpe ratio and thus determine their optimal exposures to the assets considered. However, there is an alternative criterion that has an equally plausible underlying idea; geometric mean maximization aims to maximize the growth of the capital invested, thus seeking to maximize terminal wealth. This criterion has several attractive properties and is easy to implement, and yet it does not seem to be very widely used by practitioners. The ultimate goal of this article is to explore potential empirical reasons that may explain why this is the case. The data, however, does not seem to suggest any clear answer, and, therefore, the question posed in the title remains largely unanswered: Are practitioners overlooking a useful criterion?